Amen. MS-DOS should be next, Then Windows 3.1. They are not usable for OSes today except in very limited circumstances, but would be great as learning tools. I have been saying this for years.
Wow... Babylon 5, Sliders, SeaQuest, and Voyager. These software programmers are probably facing charges on felony conspiracy to commit bad science fiction.
Thats NOT true! The first two seasons of Sliders were pretty good. Sucked like the rest after of those after season 2 tho....
Nice link, thank you for the correction. It appears my previous misconception was simply based on old data. I understand now, the "limit" exists but is not really a limit with a second or two of fancy footwork via bz, regardless of kernel size.
But if SCO do win, now the code in question is known, all that has to happen is that the code is removed, and you can now switch to the new, improved version.
You're really only going to have problems if the code can't easily be re-written, and then, it's only going to affect more enterprise-level mulit-cpu boxes - You're probably still going to be fine.
Ok, to compare: Lets say I come and burn your house down, but I pay your cost for everything inside it. You are out nothing right? Wrong.
You are out the time it takes to recover, repurchase and deal with the problem. The same here. If I have to migrate over to another kernel, or another system, it costs alot of time and money. It is not free. I am not worried that I will not be able to use any OS, its just that it will require thousands of dollars in time to deal with it. So the costs are real.
My time has value. I charge a great deal for it. It is not free if I have to change to BSD, downgrade to kernel 2.2 (which is more than a simple rebuild, since some features require it) and its not free to have to spend time thinking about it either. Since my servers are not little game machines for Counter Strike, but rather are ecommerce and support machines for an enterprize, this costs the company money as well, since they have to spend resources to correct a problem, which also affects my profit sharing and override on sales.
So yes, it affects me quite directly, in spite of so many people trying to explain how it doesn't. Ironic coming from people who don't know me.
Also, what the fool doesn't realize is it WOULD cost me money, personally, if I were to comply with the law.
If I have to either pay a royalty to SCO, or move all my servers over to a 2.2 kernel, or BSD, then it would cost more than a few thousand dollars. Also, assuming I would comply with the law, it would limit my ability to work with the 2.4+ kernels, which is something I enjoy even when not doing it for money, so it would restrict my enjoyment as well.
Your kernel has module support built in, meaning the actual kernel is less than 640k compressed, and it loads other parts of itself into upper memory. The 640k limit of x86 is still in full force. This is one reason Linux has module support, to allow a kernel larger than 640k when compressed. On all x86 systems, you must be able to load the kernel in the first 640k. after you boot that section, you can move code to any part of ram you want, but at first boot, it can only read the first 640k. Windows, DOS, OS/2, Linux, SCO, doesn't matter: its the hardware that has this limitation (hense the famous quote: 640k should be enough for anyone). Examples are modules for sound cards, video cards, etc. which are not needed at boot time so don't need to be loaded in the first 640k. Once the system initializes and can address memory beyond the first MB, then it can load those drivers/features.
For very specialized (and optimized) machines, I prefer a non modular kernel, with everything built in the primary kernel rather than as a module. Module support is smart for a desktop, and fine for servers, but I prefer a tighter kernel for dedicated tasks. It is a bit overkill perhaps, but my experience with non modular kernels is better than with them.
First, you don't need the code for loading modules in the kernel, which reduces the size of the kernel by itself. Second, everything you need to run the system is loaded at first boot, and you never need to ask the kernel to load or unload anything else, which would seem to me to be a better security model. It is also faster to build the kernel since you don't build any modules for anything else. (not that big a deal, since kernels build relatively fast on a good box anyway).
Its just a preference. When I build a box to be ONLY for DNS, for instance, there is no reason to have modules for USB, sound cards, video cards, etc. All it needs is just enough kernel to recognize the existing hardware and tcp/ip networking. This IS more secure (less likely to be affected by a new kernel bug, so you patch less often). Same reason I don't install X on any server. Just more useless software to patch, more potential security problems. If you don't absolutely need a program/feature on a server, then its a potential security problem that serves no purpose.
Part of the problem is that SCO keeps amending their case against IBM, so you are never sure what to expect next. Another problem is that what SCO says in public is NOT what they say in court. Most non technical people would serve in a jury are not really paying attention, so I can't believe its to "poisen the jury pool", as you would expect with OJ, etc. The only logical conclusion is either they really think they are right or they are artificially pumping up the stock.
I hear alot of heat about/. posting SCO stories, but this case does affect my rights online, since I use Linux mainly to run a bunch of servers. If SCO were to win, it would greatly restrict my ability to do this. This would affect my wallet and my favorite passtime, which are the same: Linux. It may not be the same as spywear, the patriot act or other topics, but it does affect my right to make a living and enjoy computing, so I would respectfully disagree with your point.
I don't know what article you read, but it sounds like SCO has a case to me.
And if I only used that one sided court filing from SCO as evidence, then I would agree. Fortunately, this is not the only piece of evidence.
First of all, IBM has taken great pains to insure that anyone on their Linux team has never had access to AIX or Sys5 code. It was setup as a clean room exercise. So, their statements are not damning, since they take great care to keep the two divisions isolated from each other.
Second, if you read something beside this ONE filing, you would know that SCO is claiming that any software or code that has ever been included in AIX is a derivative, which is not accurate. IBM had developed some journaling code in OS/2, then ported it over to AIX, yet SCO is claiming it has rights to this software, even tho it is trivial to demonstrate it is not derived from Unix, it is ported to Unix.
Third, just because a concept was introduced into Unix originally doesn't mean its infringing to be included in Linux. Some knowlege becomes virtually "Public Domain" simply because after 20 years, it has been talked about, documented, researched and experimented with by Universities and individuals. Figuring out how to impliment a feature based upon published documentation and freely available (and useable) information is called reverse engineering, not infringement. For it to be infringing, it must be a "cut and paste" job. Independent discovery is not illegal, even in America.
Fourth, IBM helping Linux is not illegal in and of itself. Implying that IBM licensed code, so any help that they were to give Linux is illegal is beyond ignorant. Their first major contribution was over 100 different printer drivers, for instance.
Fifth, IBM has extensive review about what is GPLed and what is not. They have made it perfectly clear that they will not release code wholesale, and instead are releasing code with full support and documentation, after the code has been reviewed. While they *could* make a mistake and release a portion of code that they should not have, it is unlikely considering all the checks and balances they are going through.
Sixth, there is reason to believe SCO owns the right to USE and LICENSE Unix as it sees fit, but not the copyright to actual code. There is even a lawsuit about it. But if you had read more than this one filing, you would know about Novell vs. SCO.
Seventh, So far all the code released before today as proof has been proven to be in the Public Domain and/or BSD. I have not looked at the actual code released today, but I am sure Bruce Perens will have a release within a day or three with the dirt on that. Also, some of the header files that allegedly infringe are from pre-1.0 days, and are easily documented as owned by Linus, since they were posted on USENET freely back in 91-92.
I could go on and on, but this is already more than an AC deserves.
No, x86 isn't RISC. x86 will never be RISC, although it is possible to make a RISC chip with x86 compatability. Even the new IBM 970 is more RISC than CISC, but a compromise, since RISC isn't the perfect cure-all design either. There may be RISC like features, but the instructions on the x86 are still long rather than Reduced Instructions. This is also why you can still install DOS 1.0 on a shiney new P4/3.2, or any other x86 chip, CISC and backward compatible.
I am not familiar with the book you quote, although I am familiar enough to know that the Pentium Pro, Pentium II and Pentium III all use the exact same core, with the main differences being additional instruction sets and memory/cache speeds (PPro was relatively faster but no MMX, for example). I still use primarily PPro and PIII for my servers, all dual cpu. Several IBM dual ppro 200s from 1997, still running strong with 4x2gb uwscsi40 drives and an avg. of 512M ram each.
I am waiting for the new IBM quad 970 systems to come out (2ghz x 4 for $3500, no ram, no hd) so I can get away from x86 for Linux, and be finally rid of the limitations (640k anyone?) and can more easily compile my kernels w/o module support, and more highly specialized (keep in mind I am talking about servers, not desktops here). Kinda hard to do unless you can bzImage them under 640k right now. This was easy enough in 2.2, but not necessarily so now. The most stable systems I ever had were hand rolled 2.2.x kernels, with RH 6.2 on top. Even then, it was a close fit with the x86 limits.
The AC who replied put it best. Perhaps some of the code pre-2.0 might be a bit fuzzier, but all the modern code is documented quite well. Even the SGI problem was quickly solved.
Just because SCO says that Linus doesn't know whats in the kernel, doesn't make it true.
For your information, as groklaw.net points out, SCO dropped contract claim and added copyright claim to case
I have read quite a bit of the filing, and it appears contract claims are not fully disappearing.
Quoting Groklaw:
5. This case is not about the debate about the relative merits of proprietary versus open source software. Nor is this case about IBM's right to develop and promote open source software if it decides to do so in furtherance of its independent business objectives, so long as it does so without SCO's proprietary information. This case is, and is only, about the right os SCO not to have its proprietary software misappropriated and misused in violation of its written agreements and well-settled law.
And its not over until its over. I don't know if you live in America, but as someone who worked in a law office doing paralegal and investigation, I can promise you this COULD still last a while, as far as the courts are concerned.
The judge *CAN* decide to wait until it is decided who owns the code (to potentially dismiss with prejudice). The judge can also decide to address that issue in his own court first. This is yet another contract dispute.
I agree that it doesn't look good for SCO, but it never did. They were not trying to win, they were trying to pump and dump, to inflate the stock price, and fight their way toward the door so they can take the money and run. But with the quirkiness of the courts, it still ain't over. I was being optimistic when I said this year...
That would actually be a bad thing for open source,
I agree. I admit there is a possibility that there is some bad code in Linux (ala SGI, for instance) but at least now we can look through what they are *claiming* is theirs, research the origins for that code, then make a decision. If there is any questionable code, then do the right thing: replace it. My guess is *if* there is infringing code, it would be very minor sections since any large section would have been spotted by now.
Ironic that SCO has been doing everything they can to prevent programmers from doing the right thing.
From SCO's filings: Thus, most versions of UNIX will not operate on Intel-based PC's for desktop computing; and Windows will not operate on RISC-based workstations for enterprise computing.
Perhaps the current versions don't but in the 90s, it DID run on RISC processors, WinNT 3.1, 3.5 and I think even 4.0 ran on MIPS. Since there point was relating to computing "in the 1990s" I would take their point as misleading, at the very least. What they also do not make clear is that the OLD SCO (not Caldera/SCO) was the only proprietary game for x86/unix, but even then Linux and BSD ran x86. Minor, but misleading.
I thought 2.2 was safe. Then again, we thought they were not going after copyright infringement. I'm guessing that is a typo.
Since SCO has still not actually complied with previous discovery motions, submitted millions of lines of code to IBM in paper form (real class act, they are) and keeps changing their case, my guess is we will see the end of this case, perhaps this year.
UNLESS, of course, the Novell vs. SCO suit sidetracks the IBM suit until we can figure out who actually owns Unix...
Linux is no more secure than windows against trojans
I would respectfully disagree. Linux is no more secure than windows against "social engineering", but there is a difference in a trojan run as a user and a trojan run as root. One of the primary problems with Windows is the difficulty in running some software that should be "user" software without root access.
I got my first SunOS shell many years ago, and I am pretty sure most trojans, if they had existed, might have wiped out my files, but not wiped the entire system, since I certainly did not have root access. Even at an office network, it is possible to have a Linux setup without anyone having root access, but this is more difficult with Windows, and impossible with networks that work with mixed OS's (like mine) with win9x/2k/linux.
I agree that Linux is not bullet-proof, but there are some real differences that would limit the rampant spread of a worm/trojan as long as the whole world doesn't change to Lindows or other nix varients that run as root default.
I don't think they get it. The whole idea of market cap is that market cap is exactly what the public thinks the company is worth, through the value of its stock. If the majority of people thought HP was a better/bigger/more valuable company than IBM, then their market cap would be higher. Their stock would double, IBM's stock would be halved, and the numbers would be reversed.
Market cap, determined by the stock price, is the most democratic method of determining a companies worth, because it is the opinion of every investor who buys or sells that stock, with each share having exactly the same value as another, and you can buy or sell it freely, even if you live in another country. Capitalism, in this way, is the great equalizer because it cares not for race, nationality, religion or creed.
The default for google is all languages, as it is the recommended setting. go to google and hit preferences. I makes the forever living cookie that remembers for that computer. but you can adjust several ways google acts. I usually leave it for all languages because several non english sites are still in english or have links to english versions, so its worth it as long as the results are relevent.
As to AC's, I find some of the best comments come from ACs. So do some of the worst, but the wheat is worth the chaff when it comes to ACs.
It's all about marketing. Granted descriptive names help the user when they don't have a clue, but then you can just use google or yahoo to get a clue!
But that IS marketing. Making people want your product, and using the name effectively. Granted, OSS authors don't sweat the marketability of the name as much as those of us who do marketing for a living, but that doesn't take away from the fact that the name DOES affect acceptance in a significant way.
I do think that is how it works now, it appears to be different (Ive been here a while). I think they may change the ratio to give 'receiving' points a bonus toward when you get to give them, and I usually get several points a week (capped out years ago). I metamoderate about 3 to 4 times a week, same as I visit. I find I mark alot of Troll, Offtopic and Redundant mods as Unfair. Lots of moderators get too worried about specifics that way.
Also, when I have points, I don't use user/over or the other mods that cant be metamoderated, and I almost never mod down, instead focusing on only moding good stuff up. I never moderate my own Friends or Fans, to stay unbiased. I wonder if that matters. I have been getting points twice a month for quite a while now.
GIMP: Dont want no leather clad prevert gettin medeival on my ass.
God I wish I had thought of that one, lol.
Re:Why today...
on
SCO Offline
·
· Score: 5, Interesting
Sunday isn't even a business day? How much money will they not lose?
There is one basic flaw in your assumption. Granted, for many businesses, this would hold true, but not SCO. Being attacked on Sunday is just as detrimental as being attacked on Wednesday, as it appears they make just as much money when no one is there as they do when the place is fully staffed: nothing.
I am sure they will spin this around and demonstrate how this hurt them terribly, costing them tens, if not hundreds of dollars in potential sales;) Then again, they will blame the Linux community for this, even though its soley from a bunch of owned Windows boxes. This is akin to blaming Smith and Wesson for injuries to the neighbors when you fire your gun in random directions.
yes, and if you look at the actual parent I was replying to, it was saying only 400. My guess is you have your preferences set to read at +1 only, or you would have seen the actual parent.
The reply you are referring to is the grand parent, which I was agreeing with, and explaining HOW they were different.
You might want to set your preferences to read at 0 or -1 to prevent this kind of mistake again.
Slashdot Mirror
Amen. MS-DOS should be next, Then Windows 3.1. They are not usable for OSes today except in very limited circumstances, but would be great as learning tools. I have been saying this for years.
Wow ... Babylon 5, Sliders, SeaQuest, and Voyager. These software programmers are probably facing charges on felony conspiracy to commit bad science fiction.
Thats NOT true! The first two seasons of Sliders were pretty good. Sucked like the rest after of those after season 2 tho....
Nice link, thank you for the correction. It appears my previous misconception was simply based on old data. I understand now, the "limit" exists but is not really a limit with a second or two of fancy footwork via bz, regardless of kernel size.
:D
But I still want a new IBM quad 970
But if SCO do win, now the code in question is known, all that has to happen is that the code is removed, and you can now switch to the new, improved version.
You're really only going to have problems if the code can't easily be re-written, and then, it's only going to affect more enterprise-level mulit-cpu boxes - You're probably still going to be fine.
Ok, to compare: Lets say I come and burn your house down, but I pay your cost for everything inside it. You are out nothing right? Wrong.
You are out the time it takes to recover, repurchase and deal with the problem. The same here. If I have to migrate over to another kernel, or another system, it costs alot of time and money. It is not free. I am not worried that I will not be able to use any OS, its just that it will require thousands of dollars in time to deal with it. So the costs are real.
My time has value. I charge a great deal for it. It is not free if I have to change to BSD, downgrade to kernel 2.2 (which is more than a simple rebuild, since some features require it) and its not free to have to spend time thinking about it either. Since my servers are not little game machines for Counter Strike, but rather are ecommerce and support machines for an enterprize, this costs the company money as well, since they have to spend resources to correct a problem, which also affects my profit sharing and override on sales.
So yes, it affects me quite directly, in spite of so many people trying to explain how it doesn't. Ironic coming from people who don't know me.
Also, what the fool doesn't realize is it WOULD cost me money, personally, if I were to comply with the law.
If I have to either pay a royalty to SCO, or move all my servers over to a 2.2 kernel, or BSD, then it would cost more than a few thousand dollars. Also, assuming I would comply with the law, it would limit my ability to work with the 2.4+ kernels, which is something I enjoy even when not doing it for money, so it would restrict my enjoyment as well.
Your kernel has module support built in, meaning the actual kernel is less than 640k compressed, and it loads other parts of itself into upper memory. The 640k limit of x86 is still in full force. This is one reason Linux has module support, to allow a kernel larger than 640k when compressed. On all x86 systems, you must be able to load the kernel in the first 640k. after you boot that section, you can move code to any part of ram you want, but at first boot, it can only read the first 640k. Windows, DOS, OS/2, Linux, SCO, doesn't matter: its the hardware that has this limitation (hense the famous quote: 640k should be enough for anyone). Examples are modules for sound cards, video cards, etc. which are not needed at boot time so don't need to be loaded in the first 640k. Once the system initializes and can address memory beyond the first MB, then it can load those drivers/features.
For very specialized (and optimized) machines, I prefer a non modular kernel, with everything built in the primary kernel rather than as a module. Module support is smart for a desktop, and fine for servers, but I prefer a tighter kernel for dedicated tasks. It is a bit overkill perhaps, but my experience with non modular kernels is better than with them.
First, you don't need the code for loading modules in the kernel, which reduces the size of the kernel by itself. Second, everything you need to run the system is loaded at first boot, and you never need to ask the kernel to load or unload anything else, which would seem to me to be a better security model. It is also faster to build the kernel since you don't build any modules for anything else. (not that big a deal, since kernels build relatively fast on a good box anyway).
Its just a preference. When I build a box to be ONLY for DNS, for instance, there is no reason to have modules for USB, sound cards, video cards, etc. All it needs is just enough kernel to recognize the existing hardware and tcp/ip networking. This IS more secure (less likely to be affected by a new kernel bug, so you patch less often). Same reason I don't install X on any server. Just more useless software to patch, more potential security problems. If you don't absolutely need a program/feature on a server, then its a potential security problem that serves no purpose.
Again, it is just a preference.
obviously you didnt read my entire post. try again. If you dont enable module support, it is very difficult to get the entire kernel into 640k.
Part of the problem is that SCO keeps amending their case against IBM, so you are never sure what to expect next. Another problem is that what SCO says in public is NOT what they say in court. Most non technical people would serve in a jury are not really paying attention, so I can't believe its to "poisen the jury pool", as you would expect with OJ, etc. The only logical conclusion is either they really think they are right or they are artificially pumping up the stock.
/. posting SCO stories, but this case does affect my rights online, since I use Linux mainly to run a bunch of servers. If SCO were to win, it would greatly restrict my ability to do this. This would affect my wallet and my favorite passtime, which are the same: Linux. It may not be the same as spywear, the patriot act or other topics, but it does affect my right to make a living and enjoy computing, so I would respectfully disagree with your point.
I hear alot of heat about
I don't know what article you read, but it sounds like SCO has a case to me.
And if I only used that one sided court filing from SCO as evidence, then I would agree. Fortunately, this is not the only piece of evidence.
First of all, IBM has taken great pains to insure that anyone on their Linux team has never had access to AIX or Sys5 code. It was setup as a clean room exercise. So, their statements are not damning, since they take great care to keep the two divisions isolated from each other.
Second, if you read something beside this ONE filing, you would know that SCO is claiming that any software or code that has ever been included in AIX is a derivative, which is not accurate. IBM had developed some journaling code in OS/2, then ported it over to AIX, yet SCO is claiming it has rights to this software, even tho it is trivial to demonstrate it is not derived from Unix, it is ported to Unix.
Third, just because a concept was introduced into Unix originally doesn't mean its infringing to be included in Linux. Some knowlege becomes virtually "Public Domain" simply because after 20 years, it has been talked about, documented, researched and experimented with by Universities and individuals. Figuring out how to impliment a feature based upon published documentation and freely available (and useable) information is called reverse engineering, not infringement. For it to be infringing, it must be a "cut and paste" job. Independent discovery is not illegal, even in America.
Fourth, IBM helping Linux is not illegal in and of itself. Implying that IBM licensed code, so any help that they were to give Linux is illegal is beyond ignorant. Their first major contribution was over 100 different printer drivers, for instance.
Fifth, IBM has extensive review about what is GPLed and what is not. They have made it perfectly clear that they will not release code wholesale, and instead are releasing code with full support and documentation, after the code has been reviewed. While they *could* make a mistake and release a portion of code that they should not have, it is unlikely considering all the checks and balances they are going through.
Sixth, there is reason to believe SCO owns the right to USE and LICENSE Unix as it sees fit, but not the copyright to actual code. There is even a lawsuit about it. But if you had read more than this one filing, you would know about Novell vs. SCO.
Seventh, So far all the code released before today as proof has been proven to be in the Public Domain and/or BSD. I have not looked at the actual code released today, but I am sure Bruce Perens will have a release within a day or three with the dirt on that. Also, some of the header files that allegedly infringe are from pre-1.0 days, and are easily documented as owned by Linus, since they were posted on USENET freely back in 91-92.
I could go on and on, but this is already more than an AC deserves.
No, x86 isn't RISC. x86 will never be RISC, although it is possible to make a RISC chip with x86 compatability. Even the new IBM 970 is more RISC than CISC, but a compromise, since RISC isn't the perfect cure-all design either. There may be RISC like features, but the instructions on the x86 are still long rather than Reduced Instructions. This is also why you can still install DOS 1.0 on a shiney new P4/3.2, or any other x86 chip, CISC and backward compatible.
I am not familiar with the book you quote, although I am familiar enough to know that the Pentium Pro, Pentium II and Pentium III all use the exact same core, with the main differences being additional instruction sets and memory/cache speeds (PPro was relatively faster but no MMX, for example). I still use primarily PPro and PIII for my servers, all dual cpu. Several IBM dual ppro 200s from 1997, still running strong with 4x2gb uwscsi40 drives and an avg. of 512M ram each.
I am waiting for the new IBM quad 970 systems to come out (2ghz x 4 for $3500, no ram, no hd) so I can get away from x86 for Linux, and be finally rid of the limitations (640k anyone?) and can more easily compile my kernels w/o module support, and more highly specialized (keep in mind I am talking about servers, not desktops here). Kinda hard to do unless you can bzImage them under 640k right now. This was easy enough in 2.2, but not necessarily so now. The most stable systems I ever had were hand rolled 2.2.x kernels, with RH 6.2 on top. Even then, it was a close fit with the x86 limits.
The AC who replied put it best. Perhaps some of the code pre-2.0 might be a bit fuzzier, but all the modern code is documented quite well. Even the SGI problem was quickly solved.
Just because SCO says that Linus doesn't know whats in the kernel, doesn't make it true.
For your information, as groklaw.net points out, SCO dropped contract claim and added copyright claim to case
I have read quite a bit of the filing, and it appears contract claims are not fully disappearing.
Quoting Groklaw:
5. This case is not about the debate about the relative merits of proprietary versus open source software. Nor is this case about IBM's right to develop and promote open source software if it decides to do so in furtherance of its independent business objectives, so long as it does so without SCO's proprietary information. This case is, and is only, about the right os SCO not to have its proprietary software misappropriated and misused in violation of its written agreements and well-settled law.
And its not over until its over. I don't know if you live in America, but as someone who worked in a law office doing paralegal and investigation, I can promise you this COULD still last a while, as far as the courts are concerned.
The judge *CAN* decide to wait until it is decided who owns the code (to potentially dismiss with prejudice). The judge can also decide to address that issue in his own court first. This is yet another contract dispute.
I agree that it doesn't look good for SCO, but it never did. They were not trying to win, they were trying to pump and dump, to inflate the stock price, and fight their way toward the door so they can take the money and run. But with the quirkiness of the courts, it still ain't over. I was being optimistic when I said this year...
That would actually be a bad thing for open source,
I agree. I admit there is a possibility that there is some bad code in Linux (ala SGI, for instance) but at least now we can look through what they are *claiming* is theirs, research the origins for that code, then make a decision. If there is any questionable code, then do the right thing: replace it. My guess is *if* there is infringing code, it would be very minor sections since any large section would have been spotted by now.
Ironic that SCO has been doing everything they can to prevent programmers from doing the right thing.
From SCO's filings: Thus, most versions of UNIX will not operate on Intel-based PC's for desktop computing; and Windows will not operate on RISC-based workstations for enterprise computing.
Perhaps the current versions don't but in the 90s, it DID run on RISC processors, WinNT 3.1, 3.5 and I think even 4.0 ran on MIPS. Since there point was relating to computing "in the 1990s" I would take their point as misleading, at the very least. What they also do not make clear is that the OLD SCO (not Caldera/SCO) was the only proprietary game for x86/unix, but even then Linux and BSD ran x86. Minor, but misleading.
I thought 2.2 was safe. Then again, we thought they were not going after copyright infringement. I'm guessing that is a typo.
Since SCO has still not actually complied with previous discovery motions, submitted millions of lines of code to IBM in paper form (real class act, they are) and keeps changing their case, my guess is we will see the end of this case, perhaps this year.
UNLESS, of course, the Novell vs. SCO suit sidetracks the IBM suit until we can figure out who actually owns Unix...
Linux is no more secure than windows against trojans
I would respectfully disagree. Linux is no more secure than windows against "social engineering", but there is a difference in a trojan run as a user and a trojan run as root. One of the primary problems with Windows is the difficulty in running some software that should be "user" software without root access.
I got my first SunOS shell many years ago, and I am pretty sure most trojans, if they had existed, might have wiped out my files, but not wiped the entire system, since I certainly did not have root access. Even at an office network, it is possible to have a Linux setup without anyone having root access, but this is more difficult with Windows, and impossible with networks that work with mixed OS's (like mine) with win9x/2k/linux.
I agree that Linux is not bullet-proof, but there are some real differences that would limit the rampant spread of a worm/trojan as long as the whole world doesn't change to Lindows or other nix varients that run as root default.
I don't think they get it. The whole idea of market cap is that market cap is exactly what the public thinks the company is worth, through the value of its stock. If the majority of people thought HP was a better/bigger/more valuable company than IBM, then their market cap would be higher. Their stock would double, IBM's stock would be halved, and the numbers would be reversed.
Market cap, determined by the stock price, is the most democratic method of determining a companies worth, because it is the opinion of every investor who buys or sells that stock, with each share having exactly the same value as another, and you can buy or sell it freely, even if you live in another country. Capitalism, in this way, is the great equalizer because it cares not for race, nationality, religion or creed.
So yea, I kinda agree with you.
nuc-u-lar. its pronounced "nucular".
;)
The default for google is all languages, as it is the recommended setting. go to google and hit preferences. I makes the forever living cookie that remembers for that computer. but you can adjust several ways google acts. I usually leave it for all languages because several non english sites are still in english or have links to english versions, so its worth it as long as the results are relevent.
As to AC's, I find some of the best comments come from ACs. So do some of the worst, but the wheat is worth the chaff when it comes to ACs.
It's all about marketing. Granted descriptive names help the user when they don't have a clue, but then you can just use google or yahoo to get a clue!
But that IS marketing. Making people want your product, and using the name effectively. Granted, OSS authors don't sweat the marketability of the name as much as those of us who do marketing for a living, but that doesn't take away from the fact that the name DOES affect acceptance in a significant way.
I do think that is how it works now, it appears to be different (Ive been here a while). I think they may change the ratio to give 'receiving' points a bonus toward when you get to give them, and I usually get several points a week (capped out years ago). I metamoderate about 3 to 4 times a week, same as I visit. I find I mark alot of Troll, Offtopic and Redundant mods as Unfair. Lots of moderators get too worried about specifics that way.
Also, when I have points, I don't use user/over or the other mods that cant be metamoderated, and I almost never mod down, instead focusing on only moding good stuff up. I never moderate my own Friends or Fans, to stay unbiased. I wonder if that matters. I have been getting points twice a month for quite a while now.
I think foomatic is for printing, drivers and such. I use linux for servers, but not print servers, so not positive about that tho
GIMP: Dont want no leather clad prevert gettin medeival on my ass.
God I wish I had thought of that one, lol.
Sunday isn't even a business day? How much money will they not lose?
;) Then again, they will blame the Linux community for this, even though its soley from a bunch of owned Windows boxes. This is akin to blaming Smith and Wesson for injuries to the neighbors when you fire your gun in random directions.
There is one basic flaw in your assumption. Granted, for many businesses, this would hold true, but not SCO. Being attacked on Sunday is just as detrimental as being attacked on Wednesday, as it appears they make just as much money when no one is there as they do when the place is fully staffed: nothing.
I am sure they will spin this around and demonstrate how this hurt them terribly, costing them tens, if not hundreds of dollars in potential sales
yes, and if you look at the actual parent I was replying to, it was saying only 400. My guess is you have your preferences set to read at +1 only, or you would have seen the actual parent.
The reply you are referring to is the grand parent, which I was agreeing with, and explaining HOW they were different.
You might want to set your preferences to read at 0 or -1 to prevent this kind of mistake again.